Fig. 2: The kinetics of subset HIV frequency vary by subset and are generally slower than cellular turnover.
A Longitudinal kinetics of HIV subset infection frequency in each cell subset: thin lines and dots are individual trajectories and thick solid lines represent the estimated average slopes from a log-linear mixed effects model. B Box plots of participants’ decay rates—note that some are positive, meaning that HIV frequency increased. P-values indicate one-sided t-test against null hypothesis of no clearance. For scale, the decay rate equivalent to the QVOA reservoir benchmark 44 month half-life30 is denoted with the dashed gray line. C Cellular turnover rates derived from deuterated water labeling in 24 of these 37 individuals. P values indicated paired two-sided t-tests with non-equal variance. Magnitudes of cellular turnover rates (in non-TN subsets) are much higher than HIV decay rates—note difference in y-axis scales in (C) versus (B). D The % of cellular turnover that is accompanied by HIV turnover (Methods). Values close to 100% indicate that HIV is typically repopulated when cells turn over. In (B–D) box plots indicate median (center line), interquartile range (box), 1.5x interquartile range (whiskers), and outliers (gray diamonds). Each dot (N = 24) represents an individual. E Cartoon example for TEM: in a year, there is frequent cellular turnover, which is infrequently (~5% of events) accompanied by elimination of HIV-infected cells, resulting in the observed slight decay of HIV DNA.
